According to a report by The Information, sources familiar with the matter say OpenAI is in talks with Amazon to raise as much as $10 billion—or potentially more—in additional funding. The proposed deal carries a critical condition: OpenAI would commit to using Amazon’s AI chips.
As early as November, OpenAI announced a seven-year cloud services agreement with Amazon worth up to $38 billion. Under that arrangement, AWS is set to provide OpenAI with large-scale supercomputing clusters based on Amazon EC2, with deployment expected to be completed by the end of 2026. News of further investment negotiations suggests the relationship could evolve from a straightforward customer arrangement into a deeper strategic partnership.
The most striking element of the report is the requirement that OpenAI adopt Amazon’s AI chips. Today, large-scale model training remains heavily dependent on NVIDIA GPUs, whose high prices and constrained supply have long been a bottleneck for OpenAI. In recent years, Amazon has been aggressively developing its own silicon—namely the Trainium chips for training and Inferentia chips for inference—in an effort to challenge NVIDIA’s dominance.
If OpenAI were to agree to train or run its GPT models on AWS’s in-house chips, it could significantly reduce compute costs. For Amazon, such an endorsement would serve as the most powerful validation imaginable of its chip performance. This move once again underscores a simple truth: compute is power. Despite Microsoft’s $13 billion investment to date, OpenAI is still seeking another multibillion-dollar infusion, highlighting a burn rate far beyond conventional expectations. Sam Altman, it seems, is intent on ensuring that the company’s future is not entirely beholden to a single cloud provider—or a single hardware vendor.
This approach reflects a classic risk-hedging strategy. By bringing Amazon into the fold, OpenAI gains leverage between Microsoft and Amazon. For Amazon, which missed the first explosive wave of generative AI, tying OpenAI to its own silicon through investment could provide a rare opportunity to overtake Microsoft Azure on the AI cloud battlefield.
Whether Microsoft will invoke contractual provisions to block or constrain such a deal remains an open question—and one that will be closely watched going forward.